Postgraduate Projects

The project aims to solve a widespread problem in any residential house, help residents to remember what groceries they are lacking when they are shopping in a supermarket. It has been estimated that each household are costing between $2,200 to $3,800 a year. They also added that 33% of food managers are doing a comprehensive shopping list and 27% are doing a brief one.  The idea is to have a robot arm with a camera at its end so that we can identify items present in the user pantry and save them as a database that will be copied to the phone when the user is at home. The database is updated each time the user is at home and can be consulted offline. The arm is moved with an Arduino board and 4 stepper motors, it uses a behavioural cloning algorithm with a quick inverse kinematic calculation to move the arm in a sequence that helps the camera to identify all the items. Moreover, with the camera, it takes some photos, and with object detection technology, it is able to label items present in the pantry.